DE112020002859T5 - ENCRYPTED KNOWLEDGE GRAPH - Google Patents

Abstract

A computer-implemented method for managing access rights to a knowledge graph is provided. The method includes dividing the respective portion of the knowledge graph for each user system into a plurality of knowledge subgraphs, encrypting each of the knowledge subgraphs, and creating a plurality of private summary graphs. The method also includes maintaining a collaboration graph comprising one node per user system and edges representing collaborations between the users, mapping all private knowledge subgraphs of all user systems onto a public summary graph, each node of the public summary graph having fewer data includes as the associated node of the associated private summary graph and wherein none of the nodes of the summary graph include an encryption or decryption key, and granting access to a selected knowledge subgraph from a first user system to a second user system.

Description

BACKGROUND

The invention relates generally to access rights in a knowledge management system, and more particularly to managing access rights to a knowledge graph having content managed by a plurality of entities.

One of the biggest challenges in enterprise information technology (IT) is the management of unstructured data. In the age of cognitive computing, knowledge graphs are used to store, manage, and process information outside of transactional systems. Knowledge graphs are widely accepted tools for organizing the collaboration of people within organizations and across departmental and organizational boundaries.

A knowledge graph comprises at least facts collected from a plurality of sources and typically stored in nodes of a network, and edges, links between the nodes, which represent the relationship between the nodes and thus a relationship between those stored in the knowledge graph content elements, d. H. the facts, save.

Typical knowledge graphs range in size from 1 million to 100 million nodes. Therefore, storing and managing a knowledge graph can require a significant amount of storage and computing capacity in a data center.

Knowledge graphs are typically used to facilitate collaboration within and between organizations (companies, departments and their employees). Portions of an organization's knowledge graph may include confidential data, while other portions should be made available to other selected organizations and/or users. For example, access control may be provided by a central authority as is known in the art. However, a central authority must be set up and consistently maintained. Users need to register and more importantly trust the central authority as it maintains all credentials and may have full access to all information, restricted or unrestricted. Therefore, a central authority for self-organizing organizations and decentralized responsibility for content can be prohibitive. In addition, a central authority may delay the required maintenance work and/or the granting and/or withdrawal of access rights.

EXECUTIVE SUMMARY

According to an aspect of the present invention, a computer-implemented method for managing access rights to a knowledge graph having content managed by a plurality of entities can be provided. The method may include: for each user system of a plurality of user systems, dividing its respective part of the knowledge graph into a plurality of knowledge subgraphs, encrypting each of the knowledge subgraphs using a private key of a knowledge subgraph specific asymmetric public/ private key pair, and creating a plurality of private summary graphs, one for each user system. Each private summary graph can include a node for each knowledge subgraph of the user system. In addition, each node can also include the knowledge subgraph specific asymmetric public/private key pair.

The method may further include: maintaining a collaboration graph comprising one node per user system and edges representing collaborations between the users, mapping all private subgraphs of all user systems onto a public summary graph, each node of the public summary graph comprising less data than the associated one nodes of the associated private summary graphs and wherein none of the nodes of the summary graph include any of the private keys used to encrypt any knowledge subgraph, and granting access to a selected knowledge subgraph from a first user system to a second User system by providing a decryption key related to the selected knowledge subgraph from the first user system to the second user system.

According to a further aspect of the present invention, an associated knowledge graph management system for managing access rights to a knowledge graph can be provided. The system may comprise a splitting unit capable of splitting, for each user system of a plurality of user systems, its respective part of the knowledge graph into a plurality of knowledge subgraphs, an encryption unit capable of encrypting each of the knowledge subgraphs using a pri private key of a knowledge subgraph specific asymmetric public/private key pair, and a generator capable of generating a plurality of private summary graphs, one for each user system. Each private summary graph may include a node for each knowledge subgraph of the user system, and each node may also include the knowledge subgraph-specific asymmetric public/private key pair.

In addition, the system may comprise a management unit capable of maintaining a collaboration graph comprising a node per user system and edges representing collaborations between users, and a mapping module capable of mapping all private subgraphs of all user systems to a public one to map summary graphs. Each node of the public summary graph may contain less data than the corresponding node of the corresponding private summary graph; none of the nodes of the summary graph can include any of the private keys used to encrypt any knowledge subgraph.

In addition, an access module may be adapted to grant access to a selected knowledge subgraph from a first user system to a second user system by providing a decryption key related to the selected knowledge subgraph from the first user system to the second user system.

The proposed computer-implemented method for managing access rights to a knowledge graph whose content is managed by a plurality of entities can offer numerous advantages and technical effects:

The present invention can enable a simple, straightforward, effective and efficient way to share data and knowledge through the plurality of knowledge graphs - or parts thereof - without requiring a central authentication, authorization and management authority. The knowledge graph - or the plurality of knowledge graphs - becomes, so to speak, itself manageable by the users and owners of the content stored in the one or more knowledge graphs. The decision on access rights is made directly by the content owners with their respective client systems.

This can lead to an increase in motivation to use the knowledge graph and grant or revoke access rights. This allows for much more direct control of the knowledge graph. The time processes and the administrative effort can be reduced.

Access management control to the knowledge subgraphs is distributed directly to the users. Users who "own" content in the form of a knowledge subgraph decide for themselves who to grant access to and who not to. The process of revoking access to a particular knowledge subgraph can also be in the hands of the knowledge owners.

The layer architecture - in particular the private graph layer, the collaborative graph layer and the public summary graph - are very helpful in achieving the fine-grained decentralized access control to the knowledge subgraphs.

Further embodiments of the present invention - both for the method and for the associated system - are described below:

According to an advantageous embodiment, the method may also include removing the second user system's access to the selected knowledge subgraph by re-encrypting the selected knowledge subgraph with a new private key of a knowledge subgraph-specific asymmetric public/private key pair. This allows the content owner to have complete control over their content. Even if access has been granted to another user, the content owner may revoke access to the content at any time.

According to a preferred embodiment of the method, each node of each private subgraph can include a summary description of the content of the associated knowledge subgraph. In this way, the private subgraphs can form an intermediate layer between the content owned by a particular user and the content owned by another user.

According to an expedient embodiment, the method may also include connecting the collaboration graph by creating an account for a new user system in the collaboration graph, the new user system being represented by a new node, splitting the contents of an external knowledge graph - in particular outside the main one -Wis sens graphs - into a plurality of external knowledge subgraphs, encoding the external knowledge subgraphs - as well as the main knowledge graph, and uploading the external knowledge subgraphs into the knowledge graph. In this way, new users can always decide whether they want to share their content with other users who are already using the knowledge graph of the main content. A new user can also decide which part of his content - ie the knowledge graph he manages - should be shared with other users.

According to a permitted embodiment, the method may also include searching the public summary graph by searching summary contents of the nodes of the public summary graph. The Summary Knowledge Graph is not restricted. Thus, all users can search this public area of the knowledge graph. However, this does not mean that the user has access to content related to a node in the private summary graph. You may need to be able to request access to the underlying detailed knowledge graph - i.e. the associated content graph.

According to an advantageous embodiment, the method may also include granting read access to a knowledge subgraph of a first user system - in particular after receiving an access grant request to a third user system - by using the public part of the public/private key pair of the node referring to the Knowledge subgraphs relates to which access is to be granted, provided, ie sent, from the first user system to the third user system. It should be noted that the transmission can be encrypted with the public key of the third party user system. Note also that the third user system can create/define links between the elements of its knowledge subgraph and that of the first user system, as well as add a new node in the private summary graph of the third user system. This can allow a great deal of flexibility in fine-grained content access control.

According to another advantageous embodiment, the method can also include the third user system adding edges between nodes of the knowledge subgraph of the first user system and the knowledge subgraph of the third user system. This mechanism allows each user to have the best control over their own content, especially what other users can see and read.

According to a further advantageous embodiment, the method can also include: granting write access to a knowledge subgraph of a first user system - in particular and again after receiving a corresponding write request - to a third user system by using the private/public key pair of the node that is based on relating the knowledge graph to be granted access to is provided from the first user system to the third user system (again by sending the same), and enabling a rollback option for the first user system after the third user system has entered a node - in particular has changed at least one - of a knowledge subgraph related to the first user system. In this way each user of the system can optionally also have full control over the management - in particular changing and adding new content - of the part of the knowledge graph owned by that specific user.

According to an optional embodiment of the method, the knowledge graph can be selected from a group comprising a collection of flat files, a relationship database or an object database and a graph database. Thus, basically any meaningful data organization for knowledge graphs can be used for the present invention.

According to a preferred embodiment, the method may also include enabling all user systems that have a corresponding client application installed and share a knowledge graph to perform the activity steps of sharing, encrypting, creating, maintaining, mapping and granting to perform. This set of characteristics can support appropriate management of the underlying concept of content sharing.

According to another optional embodiment of the method, the knowledge subgraphs can be partially distributed over different storage platforms. Thus, the main knowledge graph does not necessarily have to be a single consistent knowledge graph in a single storage system - but an aggregation of a plurality of knowledge subgraphs distributed across different systems and possibly also across different locations.

Furthermore, embodiments may take the form of an associated computer program pro product accessible from any computer-usable or computer-readable medium that provides program code for use by or in connection with a computer or any instruction execution system. For purposes of this specification, a computer-usable or computer-readable medium can be any device that can include means for storing, communicating, distributing, or transporting the program for use by or in connection with the instruction execution system, apparatus, or device.

character list

• 1 12 shows a block diagram of an embodiment of a computer-implemented method for managing access rights to a knowledge graph whose content is managed by a plurality of entities.
• 2 Figure 12 shows a block diagram of one embodiment of a plurality of user systems in communicative connection with a central memory.
• 3 Figure 12 shows a block diagram of one embodiment of the layered architecture of the present invention.
• 4 Figure 12 shows a diagram of one embodiment of the tier architecture with access granted to additional knowledge subgraphs.
• 5 Figure 12 shows a block diagram flowchart of the grant access and write access sub-processes.
• 6 Figure 12 shows a block diagram of a flowchart for the Access Removal process.
• 7 shows a block diagram of a system for managing access rights to a knowledge graph whose contents are managed by a plurality of entities.
• 8th Figure 12 shows a block diagram of a computer system relevant to the present invention.

DETAILED DESCRIPTION

A disadvantage of known solutions is still the central administration and the inability of known solutions to share parts of knowledge graphs with other departments and/or organizations without requiring a central authority for authentication and authorization. An administrator who has superuser rights should be made redundant and superfluous.

Therefore, it may be necessary to overcome the limitations of the known solutions and in particular to provide a knowledge management system that allows parts of a knowledge graph to be shared without a cumbersome central administration entity.

In the context of this specification, the following conventions, terms and/or expressions may be used:

The term "knowledge graph" can refer to a semantically organized data pool that relates entities - i.e. content elements - to one another. The content items can be represented as nodes that are selectively chosen by edges (or links) that define the relationship between the content items. The knowledge graph may be organized into a number of simple files as well as one or more databases.

The term "user system" may mean a computer system, typically a personal computer or workstation, capable of running software applications under the control of an operating system. The client referred to in the present invention may designate the software application. A user can be represented by such a workstation.

The term "knowledge subgraph-specific asymmetric public/private key pair" may denote a public/private key pair that is generated exclusively for a specific knowledge subgraph. The generation can be performed by the client (application). The public part can be shared while the private key is only accessible to a very limited set of systems, usually those that have access to the contents of the knowledge subgraphs.

The term "private summary graph" can denote a level in the level architecture that is organized one level above the knowledge subgraph. It can be used to organize the different knowledge subgraphs at the content level.

The term "node" can denote a logical entity that represents content or users in a knowledge graph. Nodes can also be used for administrative purposes. The nodes can be selectively chosen by edges that represent dependencies between content items, users, access rights, and/or a mixture of these.

The term "content graph" can refer to the lowest level of the content graph architecture. The knowledge graph is composed of a plurality of knowledge subgraphs that include nodes representing the content items. In the context of the present invention, the terms “content graph” and “knowledge graph” can be used synonymously; however, it can be understood that the lowest level may include a plurality of knowledge subgraphs that can be connected together to form a complete content graph, or a plurality of content graphs that are not connected together because users have not established any links between them.

The term "private summary graph" may denote a graph level above the content graph that includes a node for each knowledge subgraph of the content graph, both user and collaborator subgraphs. Basically, a private summary graph node represents a summary of the corresponding subgraphs plus additional management information (i.e. encryption/decryption keys).

The term "collaboration graph" can denote the next level of the knowledge graph above the private summary graph. Each node of the collaboration graph can represent a user of the knowledge graph. From here, links can lead to all content of all graph-level knowledge subgraphs that the user has access to, to his own knowledge subgraphs as well as to the employees' knowledge subgraphs to which he has access.

The term "public summary graph" can contain the content of the graph level - i.e. H. the summary of the knowledge subgraphs - in summarized form. The public summary graph may represent the top level of the knowledge graph architecture. A node of the public summary graph can only contain the summary of the corresponding knowledge subgraphs. The public summary graph can represent the initial level of search and navigation with unrestricted access for all users.

The term “collaboration” here can mean that one user allows another user to access the subgraphs of the content that they manage.

The term "access" may mean that a user who does not own and maintain a subgraph has allowed another user to read the content. The owning user can also be assigned to another user, e.g. B. allow an employee to write new content in the subgraphs or to change existing content, d. H. write access.

The term 're-encrypt' can mean that subgraphs - specifically the information stored in subgraphs' nodes - can be directly re-encrypted with a new key without immediately decrypting and storing them. In this way, the contents of a sub-graph can never be suspended in the time period between decryption and a new encryption. Re-encryption is the basis for revoking access to content items in relation to a user who may have previously had access to them.

A detailed description of the figures is given below. All instructions in the figures are shown schematically. First, a block diagram of an embodiment of the inventive computer-implemented method for managing access rights to a knowledge graph with content managed by a plurality of entities is shown. Further embodiments and embodiments of the knowledge graph management system for managing access rights to a knowledge graph are then described.

1 10 shows a block diagram of one embodiment of the computer-implemented method 100 for managing access rights to a knowledge graph with content managed by a plurality of entities. In general, users are represented by their computer system (ie, workstation) running an application and/or browser. The method 100 comprises dividing 102 its respective portion of the knowledge graph into a plurality of knowledge subgraphs for each user system - representing a particular user - of a plurality of user systems; encrypting 104 each of the knowledge subgraphs using a private key of a knowledge subgraph-specific asymmetric public/private key pair; and creating 106 a plurality of private summary graphs, one for each user system, each private summary graph comprising a node - specifically one - for each knowledge subgraph of the user system. Each node also includes the knowledge subgraph-specific asymmetric public/private key pair, ie both an encryption key and a decryption key.

The method 100 also includes maintaining 108 a collaboration graph comprising one node per user system and edges representing collaborations between the users; and mapping, 110, all private summary graphs of all user systems to a public summary graph, each node of the public summary graph comprising less data than - or the same amount as - the corresponding node of the corresponding private summary graph, and wherein none of the nodes of the summary graph have any includes the private keys used to encrypt any one of the knowledge subgraphs.

The method 100 also includes granting 112 access to a selected knowledge subgraph from a first user system to a second user system by providing a decryption key related to the selected knowledge subgraph from the first user system to the second user system.

2 FIG. 200 shows a block diagram 200 of an embodiment of a plurality of user systems in communicative connection (wired or wireless network) with a central memory. The underlying infrastructure thus comprises only two types of main components, the user workstations 202, 204, 206, 208 running the client (a software application) and the shared storage 210, e.g. B. a publicly accessible cloud storage. The knowledge graphs are encrypted and stored on shared memory 210 . Only the summary graph is stored unencrypted and can be accessed by all user systems. All other data is protected against access by encryption. Access must be granted individually. In one implementation, the shared storage 210 is provided as flat file storage, in another implementation as a relational database management system (RDBMS). The client (application) running on the user's workstation generates the sets of asymmetric encryption keys used, performs the encryption and decryption, and accesses the knowledge graphs.

As explained in detail below, the present invention is based on four levels of knowledge graphs. The current (formerly unencrypted) knowledge graph is called the content graph and is divided into multiple subgraphs so that specific users can access specific content. Access can be role-based. (i) The subgraphs and their encryptions are handled by (ii) the private summary graph, which is itself encrypted. (iii) The collaboration graph connects the users, represented as nodes, and transports shared encrypted keys. (iv) The unencrypted public graph allows users to search for content owned by collaborators.

By sharing keys for decryption and encryption, users can collaborate with peers, i. H. other users, d. H. other computer systems of users, d. H. other workstations to start cooperating.

3 Figure 3 shows a block diagram of one embodiment of the layered architecture 300 of the present invention. Four levels of knowledge graphs are shown: the public knowledge graph 302, the collaboration graph 304, the private summary graph 306, and the content graph 308, which includes a plurality of knowledge subgraphs. The different levels of the knowledge graphs are separated from each other by horizontal dashed lines.

The edges are stored on the links between nodes. Assuming nodes A and B are connected, then a reference to B is stored on A and a reference to A on B. In some cases, the attributes of the part edge may be different from A to B and from B to A.

An index of nodes is maintained for each knowledge graph/knowledge subgraph.

A summary of a graph includes one or more of the following: the abstraction of its structure (like a coarse graph), tags assigned to the nodes, keywords for the content of the nodes, and a list of the nodes.

Asymmetric encryption is used to protect content and make it accessible to selected individuals or groups. Asymmetric cryptography is used in such a way that users generate key pairs consisting of an encryption key K _enc (also known as a private key) and a decryption key K _dec (also known as a public key). The key generation, encryption and decryption activities are performed by the client on the user's workstation. A certification authority such as a public key infrastructure (PKI) is advantageously not required.

It should also be noted that in the following diagrams, lines without an arrowhead denote bidirectional edges and arrows denote unidirectional edges. For example, the connection from node J to the workload group 312 of the content graph 308 is unidirectional.

The lowest level contains the content graph(s) 308, i. H. the knowledge subgraph(s). Its source is the original, unencrypted knowledge graph. The knowledge graph 308 has been divided into knowledge subgraphs so that individual sets of content 310, 312, 314 (subgraphs) can be shared with peers. The owner of a particular knowledge graph can take different approaches to define subgraphs.

For example: (i) When creating or expanding the knowledge graph, the owner defines subgraphs based on content and potential collaborators. (ii) The owner uses an algorithm to find community structures and uses them to partition the knowledge graph based on its structure. Algorithms for finding communities can include the minimum method, hierarchical clustering, the Girvan-Newman algorithm, modularity maximization, statistical inference, and click-based methods. Tightly connected regions from a center of a subgraph and loosely coupled nodes are distributed to different subgraphs. (iii) The owner groups the nodes by content, e.g. B. It generates a list of topics and each node is assigned to the most appropriate topic. (iv) A combination of the above possibilities.

Each subgraph (J,K,L) 310, 312, 314 is encrypted with a different private key. By sharing the appropriate decryption key, other users can be granted access to a subgraph. The content graph may contain links to graphs owned by other users (see dashed lines 406 in 4 ).

The private summary graph 306 includes a node for each subgraph of the content graph 308 - both own subgraphs and collaborator subgraphs. Two nodes in the private summary graph 306 are connected if at least one edge exists between the nodes of the respective subgraphs. Each node of the private summary graph 306 includes: a list of the nodes of the corresponding subgraph; a summary of the corresponding subgraph; for own subgraphs and for subgraphs of collaborators with write access, the decryption and encryption key of the corresponding subgraph; for collaborator subgraphs with read-only access, the decryption key of the corresponding subgraph; and the location of the subgraph. Also note that the private summary graph is encrypted. The decryption key is initially not shared.

The collaboration graph 304 represents the connections to other users' knowledge graphs (e.g., U, C, D, B). The nodes represent the users (the user himself and the collaborators). Each user U, C, D, B is assigned a unique identifier that is generated when the user joins the system. The edges between the nodes of the collaboration graph summarize the edges that arise between the nodes of the user subgraph(s) and the collaborator subgraph(s). If multiple collaborators grant access to their subgraph(s) and edges between their subgraphs have been found, the link between the collaborators' subgraphs is merged into the collaboration graph (e.g., between nodes B and C).

A node in collaboration graph 304 that represents the user (e.g., U, C, D, B) is linked to all nodes owned by the user in both public summary graph 302 and private summary graph 306. A collaborator can, in principle, follow an edge to a node in the private summary graph 306, but has no access to the node (and its subsequent edges) since the node is encrypted. A user node also receives and stores access requests from collaborators.

Public summary graph 302 (like private summary graph 306) includes a node (e.g., J, K, L, N, O) for each subgraph. However, a node only includes the summary of the corresponding subgraph. It does not include keys or a list of nodes. The summary in the public summary graph 302 may not contain as much information as in the private summary graph. However, it should be comprehensive enough to provide a meaningful search result.

A node also contains the location of the subgraph in the content graph. Each node is linked to the user node of the node's owner. The client only adds an edge between the nodes of different owners if a (at least one-sided) collaboration has come about and at least one edge added between the corresponding subgraphs of the content graph.

In one implementation, the central knowledge graph repository (see 2 , 210) implemented as a set of simple files, where a node is represented by a file and the edges are stored in the node file as references to other files (ie path and filename).

In another implementation, an industry standard knowledge graph database is used. These graph databases are relational databases in which each object (a node) has a primary, unique key. Other objects related to an object (nodes connected by edges) are referenced by "foreign keys", which are the values of primary keys stored as an attribute in the object.

In one implementation, a single graph database is used by all clients. In another implementation, multiple graph databases are used, or even each client sets up its own graph database. Of course, all graph databases must be compatible, i. H. follow the same standard and each graph database must have a unique identifier. All databases allow (at least read) access via a generic user ID (identifier) used by the clients, i.e. the applications on the users' workstations.

Note that the public summary graph 302 is disjoint; that is, two or more subgraphs have no edges in common. This is at least the case when a new user joins (see below). After connected and shared subgraphs are created, respective users can define edges between the subgraphs.

If a single graph database is used, the client accesses the table of nodes of the public summary graph 302, which includes a pointer to the content for each node (referred to above as "location").

If multiple graph databases are used, the client searches the storage system for graph databases and accesses the table of public summary graph nodes used in each database, which contains a pointer to the contents for each node (referred to above as "location") . When multiple graph databases are used, the primary keys are not unique across databases; hence the edges are implemented as tuples (database identifier, primary key).

Common unit operations are: joining users, searching the summary graph, granting access, revoking or removing access. These different scenarios are described below.

Joining a User

If a new user wants to join the collaboration network, he or she must sign up for shared storage. Each user must have a client installed on their workstation that supports access to and maintains a knowledge graph.

First, the user creates a node that represents them in the collaboration graph 304 (exemplified as user A in 3 ). The node that represents the user is initially not linked to any other node.

The user decides the number of initial subgraphs. The user creates one or more key pairs (Key ⁱ _enc , Key ⁱ _dec ), one for each subgraph, e.g. i=J, K, L, as in 3 shown. On the workstation, the user creates the subgraphs either by adding nodes and edges to them or by importing an existing knowledge graph and breaking it up into subgraphs as described above. The client encrypts each subgraph with the appropriate key Key ⁱ _enc and uploads the subgraph to the central store. Because the edges are stored along with the nodes, edges connecting subgraphs are found in both subgraphs.

For each subgraph uploaded, the client adds a node to the summary graph and adds the edges to that node as described above.

The user's client generates the "user-specific" key pair Key ^A _enc and Key ^A _dec and encrypts the private summary graph with this key. Key ^A _dec is only stored on the user's workstation and is never shared. Key ^A _enc is stored on the collaboration graph and is accessible to (potential) collaborators.

Browsing the summary graph

Any graph search algorithm may be used to search the public summary graph 302 . In one implementation, a sparsely connected graph algorithm is used. In another implementation, the table(s) of the graph database listing the nodes of the summary graph J,K,L,N,O can be traversed in a known manner.

Granting Access

4 FIG. 4 shows an embodiment of the tier architecture 400 with access granted access to additional knowledge subgraphs. 4 describes the process of granting access to another user's content. 5 Figure 5 shows a block diagram of flowchart 500 of the grant access and write access sub-processes.

After a first user - e.g. user A - has searched and recognized 502 content of interest in one or more subgraphs owned by a second user - e.g. user B -, requests 504 user A access to these subgraph(s) - e.g. subgraphs 402, 404- on. In this process, user A's client stores an access request on user B's node in the collaboration graph 306. User B's client fetches the request.

User B receives 506 notification and can grant (or deny) access. Following a decision 508 by User B to grant access, User B's client retrieves 510 the decryption key(s) of the content subgraph(s) (e.g., 402, 404, 4 ) from the corresponding nodes of private summary graphs N, O ab, encrypts 512, then those private summary graph decryption keys 306 with user A's encryption key (retrieved from the collaborative graph) and stores them on user A's node in the collaborative graph 304 .

To illustrate this process, 4 , showing a view centered on user A, expanded to include the details for user B, represented by dashed lines from node N of left private summary graph 306 to subgraph 402 and from node O of left private summary graph 306 (surrounded by dashed lines) to subgraph 404 are indicated.

Thus, user A has searched the public summary graph 302 and found interesting content at nodes N and O. Note that user A started the search either at node N or O or at a node connected by a path to N or O, since there is no (direct) connection between L and N at that point.

User A's client recognizes nodes N and O as owned by User B and creates an access request to User B's node. User B grants access, 514. User B's client gets User A's encryption key (Key ^A _enc ) from the collaboration graph 304 and two decryption keys Key ^N _dec and Key ^O _dec from its own private summary graph 306 (right part of this graph level). Using Key ^A _enc , User B's client encrypts the Key ^N _dec and Key ^O _dec and stores them 516 on User A's node in the collaboration graph 308.

User A's client retrieves the two encrypted keys from User A's nodes and decrypts them using Key ^A _dec . In the workstation's memory, the client creates two summary nodes N and O by copying the public summary graph nodes, storing Key ^N _dec and Key ^O _dec accordingly, and (optionally) enriching the two new nodes N and O by creating subgraphs 402 , 404 relating to nodes N and O with an algorithm preselected by user A. Finally, the client encrypts the two new summary nodes with key ^A _enc and adds them to user A's private summary graph 306 (right part of knowledge subgraph level).

write access

Changing nodes or edges of an encrypted knowledge graph involves including the encryption in the writing process. If storage is in flat files (each file represents a node and its edges), the user only needs to encrypt one file per node update. If a knowledge graph database is implemented, column level or (preferably) field level encryption is used to minimize the computational cost of a node update.

When working in a multi-user environment, it is important to remember that encryption does not prevent users from overwriting or copying data with potentially malicious intent Clear. This procedure supports two approaches:

Approach A) In one implementation, the store can grant unrestricted write access to all users. This means that (1) a collaborator who has received the appropriate encryption key can update content in such a way that the owner cannot distinguish which of the collaborators updated a node when multiple collaborators have received write access, and (2) any user Can delete content regardless of ownership of a key.

The following procedure is provided for this implementation. A hash key is calculated for each node (including its edges) after a write operation. Note that deletion results in hash key 0. The owner maintains a register of all hash keys of its subgraphs and a backup of all subgraphs. These can be stored either on the user's workstation or on a separate central storage system (not shown). At a predefined interval, the client checks the current hash key and, in the event of a discrepancy, informs the user about a write or delete operation on a subgraph. The user can opt for a rollback operation.

From the point of view of the process, requesting and granting write access works the same as requesting and granting read access, but in addition to the decryption keys and the encryption keys for the in-scope subgraphs, the keys (in encrypted format) must go to the requestor's collaboration graph and in turn to the private Summary graphs of the requester are added.

This approach has the advantage that changes can be made in real time by invited collaborators.

Approach B) In another implementation, the creator of a subgraph is only allowed to change a subgraph, specifically the data object containing the subgraph. If storage is implemented as simple files, the data object can be a directory. If a knowledge graph database is implemented, the data object is a database table.

Initially when a user joins the system, the client creates a user ID. The user ID is stored on the workstation. Whenever a new subgraph/data object is created, this user identifier is stored as metadata with the data object. The client sends the user ID with each write request. When the storage system receives a request to write to the data object, the user ID is compared to the metadata of the data object and only if there is a match does the storage system perform the write operation. Updates by collaborators are transported (via the owner's user node) as changed nodes to the subgraph's owner, i. H. either as a simple file or as a database table row. The owner's client indicates the change to the owner in conjunction with the surrounding knowledge graph, and the owner decides whether to accept or reject the change. For trusted collaborators, the owner can choose to automatically accept the changes. Upon acceptance, the owner's client appends the changes to the subgraph, encrypted with the subgraph's corresponding key.

The advantage of this approach is that the owner controls changes before they take effect. Thus, a higher degree of content consistency can be achieved in a collaborative environment.

Removing Access

The removal of access is performed by re-encryption, ie content encrypted with Key ^X _enc that could previously be decrypted with Key ^X _dec is processed in such a way that it can only be decrypted with the new Key ^Y _dec . A combination of Key ^X _dec and Key ^Y _enc is used for re-encryption.

In the prior art, re-encryption is mostly discussed in connection with proxy re-encryption (PRE). In the present invention, a combined key is generated to encrypt a subgraph directly, i. H. without first decrypting it and then encrypting it, d. H. without suspending the subgraph in the time period between the completed decryption and the beginning of the encryption.

6 Figure 12 shows a block diagram of a flowchart of the Remove Access sub-process. Following the previous example, user B wants to remove user A's access 602 to subgraph N only. User B's client creates 604 a new key pair (Key ^M _dec , Key ^M _enc ) and a combined key [Key ^N _dec | _{Key Menc} ^] . The client encrypts 606, the subgraph N with key [Key ^N _dec | Key ^Menc ] and stores 608 a notification of the removal of access to user A's node in the collaboration graph. User A's client removes 610 the summary node from user A's private summary graph.

In case user B has shared subgraph N with other users (e.g. C and D) and wants them to continue having access, user B's client also stores a re-encryption notification on the nodes from user C and user D along with the new decryption key Key ^M _dec encrypted with Key ^C _enc and with Key ^D _{enc respectively} to ensure that only users C and D can use Key ^M _dec .

Browsing a graph

It should be noted that unencrypted data associated with a private summary graph 306 or content graph 308 is stored only on users' workstations. Buffering a portion of a graph on the workstations can improve performance, and typically the larger the buffer, the better the performance. However, the method does not require buffering of any particular size.

A graph search starts at a given node (or a set of nodes in parallel search) and traverses the graph following the edges. At the beginning of the search, the given node is decrypted (if it is not already unencrypted in the client's buffer). While the search process is processing the first node, all nodes connected to the first node are decrypted in parallel. In turn, while the search processes the contents of the nth node, all nodes connected to the nth node are decrypted. If the client buffer is full, the nodes at the beginning of the search path are removed from the buffer.

If the search path crosses a subgraph boundary, either by entering another subgraph of the same user or a collaborator, the decryption key is replaced. In one implementation, the process checks beforehand in the private summary graph whether the current decryption key is valid for the subsequent node. In another implementation, the search process retrieves an encryption key from the summary graph only if a node's decryption failed due to the use of an incorrect key.

If the process enters a subgraph for which no decryption key is available, e.g. e.g. because the collaborator revoked access (i.e. re-encrypted the subgraph), the search algorithm considers the edge to that subgraph non-existent and redirects the search path.

Managing knowledge subgraphs

If the owner of a subgraph only wants to share parts of the subgraph with a collaborator, he must split the subgraph. The client decrypts the subgraph at the workstation, splits the subgraph using the selected method, encrypts the new set of subgraphs with new, different keys, replaces each subgraph with the relevant set of subgraphs, and updates the public summary graph and the private summary graph of the owner. In order for the collaborators to retain their access, the client encrypts the new set of subgraph keys with the collaborators' keys and stores them on the respective users' nodes. Upon addition, a message related to the updated public summary graph is stored on the users' nodes, so that each collaborator's client can update the private summary graph accordingly.

If the owner of a subgraph wants to combine several subgraphs into one subgraph, he selects the subgraphs in the client. The client determines the largest of the subgraphs either in terms of the number of nodes or in terms of the total memory used by the nodes. The client decrypts the other n-1 subgraphs on the workstations, encrypts them with the key of the largest subgraph, replaces the subgraphs of the newly encrypted subgraphs, and updates the owner's public summary graph and private summary graph. For the collaborators that previously had access to at least the largest subgraph, the client stores a message at the collaborators' nodes to inform them of the summary graph update. For collaborators who previously had access to other subgraphs but not the largest subgraph, the client stores the personally encrypted subgraph encryption key and a message to the collaborators' nodes to notify them of the summary graph update.

Alternatively, the owner might decide to not access collaborators first all subgraphs had to not grant access to consolidated graphs.

7 FIG. 7 shows a block diagram of the knowledge graph management system 700 for managing access rights to a knowledge graph that is stored in one or more storage units and the content of which is managed by a plurality of entities. The system 700 comprises a dividing unit 702 which is suitable for dividing, for each user system of a plurality of user systems, its respective part of the knowledge graph into a plurality of knowledge subgraphs (see FIG 3 , 312, 312, 314), an encryption unit 704 adapted to encrypt each of the knowledge subgraphs using a private key of a knowledge subgraph-specific asymmetric public/private key pair, and a generator 706 adapted to generate a create multiple private summary graphs, one for each user system. Each private summary graph (see 3 , 306) a node for each knowledge subgraph of the user system, each node also comprising the knowledge subgraph specific asymmetric public/private key pair.

The system 700 also comprises a management unit 708 suitable for maintaining a collaboration graph comprising one node per user system and edges representing collaborations between the users; a mapping module 710 adapted to map all private summary graphs of all user systems to a public summary graph, each node of the public summary graph comprising less data than the corresponding node of the corresponding private summary graph, and wherein none of the node of the summary graph comprises any of the private keys , which are used to encode any one of the knowledge subgraphs; and an access module 712 adapted to grant access to a selected knowledge subgraph from a first user system to a second user system by sending a decryption key related to the selected knowledge subgraph from the first user system to the second user system provides.

Embodiments of the invention may be implemented in conjunction with virtually any type of computer, regardless of platform, suitable for storing and/or executing program code. 8th shows a computer system 800 as an example, which is suitable for the execution of program code in connection with the proposed method. Each workstation 202 (see 2 ) may be implemented in the form of computer system 800, as well as any system that includes central storage system 204 (see 2 ) and controls the nodes and servers that control the knowledge graphs and knowledge subgraphs.

Computer system 800 is only one example of a suitable computer system and is not intended to suggest any limitation as to the scope or functionality of embodiments of the invention described herein, regardless of whether computer system 800 is capable of implementing any of the above functions and/or to execute. Computer system 800 includes components that are compatible with numerous other general or specific computer system environments or configurations. Examples of known computing systems, environments, and/or configurations that may be suitable for use with computer system/server 800 include personal computer systems, server computer systems, thin clients, thick clients, handheld or laptop devices, multiprocessor systems, among others , microprocessor-based systems, set-top boxes, programmable consumer electronics, network personal computers, minicomputer systems, mainframe systems, and distributed cloud computing environments incorporating any of the foregoing systems or devices, and the like. Computer system/server 800 may be described in the general context of instructions executed by computer system 800, such as program modules. In general, program modules can contain routines, programs, objects, components, logic, data structures, etc. that perform specific tasks or implement specific abstract data types. Computer system/server 800 may be deployed in distributed cloud computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed cloud computing environment, the program modules can reside in both local and remote storage media of the computer system, including storage devices.

As in 8th As illustrated, computer system/server 800 is illustrated in the form of a general purpose computing device. The components of computer system/server 800 may include, but are not limited to, one or more processors or processing units 802, system memory 804, and a bus 806 that connects various system components including system memory 804 to processor 802. Bus 806 represents one or more of several types of bus structures, including a memory bus or memory controllers, a peripheral bus, an accelerated graphics port, and a processor or local bus using one of several bus architectures. Illustrative and non-limiting examples of such architectures are the Industry Standard Architecture (ISA) bus, the Micro Channel Architecture (MCA) bus, the Enhanced ISA (EISA) bus, the Video Electronics Standards Association (VESA) local bus, and the PCI bus (Peripheral Component Interconnects). Computer system/server 800 typically includes a variety of computer system-readable media. This media can be any available media that can be accessed by computer system/server 800 and includes both volatile and non-volatile media, removable and non-removable media.

System memory 804 may include computer system-readable media in the form of volatile memory, such as RAM 808 and/or cache memory 810. Computer system/server 800 may further include other removable/non-removable, volatile/non-volatile computer system storage media. As an example only, a storage system 812 for reading from and writing to a non-removable, non-volatile magnetic medium (not shown and typically referred to as a "hard drive") may be provided. Although not shown, a magnetic disk drive for reading from and writing to a removable, non-volatile magnetic disk (e.g., a "floppy disk") and an optical disk drive for reading from or writing to a removable, non-volatile optical disk such as a CD-ROM, DVD-ROM or other optical media can be provided. In such cases, each drive may be connected to bus 806 through one or more disk interfaces. As further illustrated and described below, memory 804 may include at least one program product having a number (e.g., at least one) of program modules configured to perform the functions of embodiments of the invention.

By way of example and not limitation, the program/utility having a set (at least one) of program modules 816 may be stored in memory 804, as well as an operating system, one or more application programs, other program modules, and program data. Each of the operating systems, one or more application programs, other program modules, and program data, or a combination thereof, may include an implementation of a network environment. The program modules 816 generally perform the functions and/or methods of embodiments of the invention described herein.

Computer system/server 800 may also communicate with one or more external devices 818, such as a keyboard, pointing device, display 820, etc.; with one or more devices that enable a user to interact with computer system/server 800; and/or any device (e.g., network card, modem, etc.) that enables computer system/server 800 to communicate with one or more other computing devices. This communication can be done via the input/output (I/O) interfaces 814. However, computer system/server 800 may communicate with one or more networks, such as a local area network (LAN), a general wide area network (WAN), and/or a public network (e.g., the Internet), via network adapter 822. As illustrated, network adapter 822 can communicate with the other components of computer system/server 800 via bus 806 . It is understood that although not shown, other hardware and/or software components may be used in connection with computer system/server 800. Examples include but are not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives and data archival storage systems, etc.

In addition, the knowledge graph management system 700 can be connected to the bus 806 to manage access rights to a knowledge graph - and in particular to parts thereof.

The descriptions of the various embodiments of the present invention have been presented for purposes of illustration but are not intended to be exhaustive or limited to the disclosed embodiments. Many modifications and changes will become apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles of the embodiments, practical application, or technical improvement over technologies found in the marketplace, or to enable others skilled in the art to understand the embodiments disclosed herein.

The present invention can be a system, a method and/or a computer program product. The computer program product may comprise a computer-readable storage medium (or media) having computer-readable program instructions embodied thereon causing a processor to carry out aspects of the present invention.

The computer-readable storage medium may be a tangible device that can store and maintain instructions for use by an instruction execution device. The computer-readable storage medium may be, for example but not limited to, an electronic storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. A non-exhaustive list of more specific examples of a computer-readable storage medium includes the following: a portable computer floppy disk, a hard disk, random-access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM, or flash memory), static random access memory (SRAM), portable compact disc random access memory (CD-ROM), digital versatile disc (DVD), memory stick, floppy disk, mechanically encoded device such as punch cards or raised structures in a groove on which instructions are recorded, and any suitable combination of the foregoing. A computer-readable storage medium, as used herein, is not intended to be a transient signal in itself, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating in a waveguide or other transmission media (e.g., pulses of light traveling through a fiber optic cable running), or electrical signals transmitted over a wire.

The computer-readable program instructions described herein may be downloaded to appropriate computing and processing devices from a computer-readable storage medium or over a network, such as the Internet, a local area network, a wide area network, and/or a wireless network. The network may include copper transmission cables, optical transmission fibers, wireless transmission, routers, firewalls, switches, gateway computers, and/or edge servers. A network adapter card or network interface in each computing/processing device receives computer-readable program instructions from the network and forwards the computer-readable program instructions for storage on a computer-readable storage medium within the respective computing/processing device.

Computer readable program instructions for performing operations of the present invention may be assembly language instructions, instruction set architecture (ISA) instructions, machine instructions, machine dependent instructions, microcode, firmware instructions, state setting data, or either source code or object code, in any combination of one or more programming languages, including an object-oriented programming language such as Smalltalk, C++ or the like, and conventional procedural programming languages such as the "C" programming language or similar programming languages. The computer-readable program instructions may execute entirely on a user's computer, partially on the user's computer, as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on the remote computer or server. In the latter case, the remote computer can be connected to the user's computer over any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (e.g., over the Internet using an internet service provider). In some embodiments, electronic circuits, including, for example, programmable logic circuits, field-programmable gate arrays (FPGA), or programmable logic arrays (PLA), that execute computer-readable program instructions by using the status information of the computer-readable program instructions to personalize the electronic circuit to perform aspects of the present invention.

Aspects of the present invention are described herein with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It is understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer readable program instructions.

These computer-readable program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable computing device to produce a machine, such that the instructions, executed via the processor of the computer or other programmable computing device, provide means for implementing the flowchart shown - and/or create block diagram blocks or functions/actions specified in the block diagrams. These computer-readable program instructions can also be stored on a computer-readable storage medium that includes a computer, a programmable data processing device device and/or other devices to function in a particular manner such that the computer-readable storage medium having the instructions stored thereon comprises an article of manufacture containing instructions that perform aspects of the function specified in the flowchart and/or block diagram block or blocks /Implement plot.

The computer-readable program instructions may also be loaded onto a computer, other programmable computing device, or other device to cause a series of operational steps to be performed on the computer, other programmable device, or other device to create a computer-implemented process such that the instructions executed on the computer, other programmable device, or other device implement the functions/acts specified in the flowchart and/or block diagram block or block diagrams.

The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which includes one or more executable instructions for implementing the specified logical function(s). It should be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks presented sequentially may execute substantially simultaneously, or the blocks may sometimes execute in reverse order, depending on the functionality involved. It is also noted that each block of the block diagrams and/or illustrative flowcharts, and combinations of blocks in the block diagrams and/or illustrative flowcharts, may be implemented by special purpose hardware based systems that perform the specified functions or acts or combinations of special purpose Run hardware and computer commands.

The descriptions of the various embodiments of the present invention have been presented for purposes of illustration, but are not intended to be exhaustive or limited to the embodiments disclosed. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the invention. The terminology used herein was chosen to best explain the principles of the embodiment, practical application, or technical improvement over existing technologies, or to enable those of ordinary skill in the art to understand the embodiments disclosed herein.

Claims (20)

Computer-implemented method for managing access rights to a knowledge graph with content managed by a plurality of user systems, the computer-implemented method comprising: dividing, by one or more processors, a respective portion of the knowledge graph into a plurality of knowledge subgraphs for each user system of the plurality of user systems; encrypting each of the knowledge subgraphs by one or more processors using a private key of a public/private key pair; generating, by one or more processors, a plurality of private summary graphs, wherein: a private summary graph is generated for each user system, each private summary graph comprises a node for each knowledge subgraph of the respective user system, and each node includes the public/private key pair; maintaining, by one or more processors, a collaboration graph comprising a node for each user system and edges representing collaborations between the plurality of user systems; mapping, by one or more processors, the plurality of private summary graphs of the plurality of user systems to a public summary graph; and granting access to a selected knowledge subgraph from a first user system to a second user system by providing a decryption key related to the selected knowledge subgraph from the first user system to the second user system by one or more processors. Computer-implemented method claim 1 further comprising: revoking, by one or more processors, access to the selected knowledge subgraph for the second user system by how the encryption of the selected knowledge subgraph with a second private key of a second key pair. Computer-implemented method claim 1 further comprising: connecting the collaboration graph by one or more processors by: creating, by one or more processors, an account for a new user system in the collaboration graph, the new user system being represented by a new node, splitting contents of a second knowledge graphs into a plurality of second knowledge subgraphs by one or more processors, encoding by one or more processors the plurality of second knowledge subgraphs, and uploading by one or more processors the plurality of second knowledge subgraphs into the knowledge graph . Computer-implemented method claim 1 further comprising: granting read access to a knowledge subgraph of a third user system to a fourth user system by providing, by one or more processors, a public portion of a node's public/private key pair related to the knowledge graph. Computer-implemented method claim 4 further comprising: the fourth user system adding edges between nodes of the knowledge subgraph of the third user system and the knowledge subgraph of the fourth user system by one or more processors. The computer-implemented method claim 1 further comprising: granting, by one or more processors, write access to a knowledge subgraph of a third user system to a fourth user system by accessing the third user system's private/public key pair of the node related to the knowledge graph to be granted to the fourth user system; and enabling, by one or more processors, a rollback option for the third user system when the fourth user system has changed a node of a knowledge subgraph related to the third user system. Computer-implemented method claim 1 further comprising: enabling, by one or more processors, the plurality of user systems that collectively access the knowledge graph to perform the splitting, encrypting, generating, maintaining, mapping, and granting steps. Knowledge graph management system for managing access rights to a knowledge graph stored in one or more storage units with content managed by a plurality of user systems, the knowledge graph management system comprising: a splitting unit adapted to split a respective part of the knowledge graph into a plurality of knowledge subgraphs for each user system of the plurality of user systems; an encryption unit adapted to encrypt each of the knowledge subgraphs using a private key of a public/private key pair; a generator suitable for generating a plurality of private summary graphs, wherein: a private summary graph is generated for each user system, each private summary graph comprises a node for each knowledge subgraph of the respective user system, and each node comprises the public/private key pair; a manager adapted to maintain a collaboration graph comprising a node for each user system and edges representing collaborations between the plurality of user systems; a mapping module operable to map the plurality of private summary graphs of the plurality of user systems to a public summary graph; and an access module adapted to grant access to a selected knowledge subgraph from a first user system to a second user system by providing a decryption key related to the selected knowledge subgraph from the first user system to the second user system. Knowledge Graph Management System claim 8 , wherein the access module is further adapted to remove the second user system's access to the selected knowledge subgraph by re-encrypting the selected knowledge subgraph with a second private key of a second public/private key pair. The Knowledge Graph Management System claim 8 , further comprising: a connection unit adapted to connect the collaboration graph by: creating an account for a new user system tem in the collaboration graph, wherein the new user system is represented by a new node, splitting contents of a second knowledge graph into a plurality of second knowledge subgraphs, encrypting the plurality of second knowledge subgraphs, and uploading the plurality of second knowledge Subgraphs in the knowledge graphs. Knowledge Graph Management System claim 8 wherein the access module is further adapted to grant a fourth user system read access to a knowledge subgraph of a third user system by providing a public part of a node's public-private key pair relating to the knowledge graph. system after claim 11 further comprising a maintenance module adapted for the fourth user system to add edges between nodes of the knowledge subgraph of the third user system and the knowledge subgraph of the fourth user system. Knowledge Graph Management System claim 8 wherein the access module is further adapted for: granting write access to a knowledge subgraph of a first user to a third user system by providing the node's private/public key pair related to the knowledge graph to access from the first user system the third user system is to be granted; and enabling a rollback option for the first user system when the third user system has changed a node of a knowledge subgraph related to the first user system. Knowledge Graph Management System claim 8 wherein the plurality of user systems are further adapted to share access to the knowledge graph to perform the splitting, encrypting, generating, maintaining, mapping and granting steps. Computer program product for managing access rights to a knowledge graph with content managed by a plurality of user systems, the computer program product comprising: one or more computer-readable storage media and program instructions stored on the one or more computer-readable storage media, the program instructions comprising: program instructions for dividing a respective portion of the knowledge graph into a plurality of knowledge subgraphs for each user system of the plurality of user systems; program instructions for encrypting each of the knowledge subgraphs using a private key of a public/private key pair; Program instructions for creating a plurality of private summary graphs, where: a private summary graph is generated for each user system, each private summary graph comprises a node for each knowledge subgraph of the respective user system, and each node includes the public/private key pair; program instructions for maintaining a collaboration graph comprising a node for each user system and edges representing collaborations between the plurality of user systems; program instructions for mapping the plurality of private summary graphs of the plurality of user systems to a public summary graph; and Program instructions for granting access to a selected knowledge subgraph from a first user system to a second user system by providing a decryption key related to the selected knowledge subgraph from the first user system to the second user system. computer program product claim 15 further comprising: program instructions for revoking access to the selected knowledge subgraph for the second user system by re-encrypting the selected knowledge subgraph with a second private key of a second key pair. computer program product claim 15 , further comprising: program instructions for connecting the collaboration graph by: program instructions for creating an account for a new user system in the collaboration graph, the new user system being represented by a new node, program instructions for dividing contents of a second knowledge graph into a plurality of second knowledge subgraphs, program instructions for encrypting the plurality of second knowledge subgraphs, and program instructions for uploading the plurality of second knowledge subgraphs into the knowledge graph. computer program product claim 15 further comprising: program instructions for granting read access to a knowledge subgraph of a third user system to a fourth user system by providing a public portion of a node's public/private key pair related to the knowledge graph. The computer program product after Claim 18 further comprising: program instructions for the fourth user system to add edges between nodes of the knowledge subgraph of the third user system and the knowledge subgraph of the fourth user system. computer program product claim 15 further comprising: program instructions for granting write access to a knowledge subgraph of a first third system to a fourth user system by providing the private/public key pair of the node related to the knowledge graph upon access from the third user system the fourth user system is to be granted; and program instructions for enabling a rollback option for the third user system when the fourth user system has changed a node of a knowledge subgraph related to the third user system.

Images